Design of Ionic Liquid as Carbon Capture Solvent ... - ACS Publications

Research Article pubs.acs.org/journal/ascecg

Design of Ionic Liquid as Carbon Capture Solvent for a Bioenergy System: Integration of Bioenergy and Carbon Capture Systems Fah Keen Chong,† Viknesh Andiappan,‡ Denny K. S. Ng,† Dominic C. Y. Foo,† Fadwa T. Eljack,§ Mert Atilhan,∥ and Nishanth G. Chemmangattuvalappil*,† †

Department of Chemical and Environmental Engineering/Centre of Excellence for Green Technologies, University of Nottingham, Malaysia Campus, Broga Road, 43500 Semenyih, Selangor, Malaysia ‡ Energy and Environmental Research Group, School of Engineering, Taylor’s University, Lakeside Campus, No. 1 Jalan Taylor’s, 47500 Subang Jaya, Selangor, Malaysia § Department of Chemical Engineering, College of Engineering, Qatar University, P.O. Box 2713, Doha, Qatar ∥ Department of Chemical Engineering, Texas A&M University at Qatar, Doha, Qatar S Supporting Information *

ABSTRACT: Current atmospheric carbon dioxide (CO2) concentration has exceeded the safe limit of 350 ppm. One potential technology to remove CO2 from the atmosphere is the integrated bioenergy production and carbon capture system. A bioenergy production system produces multiple energy products from biomass, resulting in zero net increment of CO2 amount in the atmosphere. Meanwhile, CO2 produced from bioenergy production is separated for other purposes through carbon capture. To ensure the entire system is environmental friendly, an efficient and green carbon capture solvent should be utilized. Ionic liquids (ILs) are the potential solvents for this purpose, as they have negligible vapor pressure and high thermal stability. However, there are up to a million possible combinations of cations and anions that may make up ILs. This work presents a systematic approach to identify an optimal IL solvent to separate CO2 produced from a bioenergy system at the optimal conditions of carbon capture process. Following that, the bioenergy system is retrofitted to provide sufficient utilities to a carbon capture system to make sure that the entire process is self-sustainable. A case study involving an existing palm-based bioenergy system, integrated with carbon capture to produce CO2, is used to demonstrate the presented approach. KEYWORDS: Bioenergy production system, Computer-aided molecular design, Decision-making tool, Disjunctive programming, Input−output modeling

■

INTRODUCTION

result in negative CO2 emissions; this technology yields zero net CO2 and also removes CO2 from the atmosphere.5 Currently, the most established CO2 separation technology is chemical absorption using an amine-based solvent. However, amines show some drawbacks such as solvent loss, solvent degradation,6 and high energy consumption during solvent regeneration.7 To overcome the aforementioned problems, ionic liquids (ILs) have been introduced as a potential replacement of amine-based solvents.8 ILs are salts that remain in a liquid state at around room temperature, consisting largely of organic cations and anions (can be either organic or inorganic).9 ILs have a wide range of thermophysical properties, and they can be customized to meet target properties for specific purposes,10 including replacement of amine-based solvents for carbon capture purposes. ILs were first considered as alternative solvents of amines by their CO2 sorption

Human activities, such as combustion of fossil fuels for energy production, have altered the carbon cycle by adding more carbon dioxide (CO2) to the atmosphere.1 The increase in CO2 levels prompted significant attention on seeking alternative means for energy production, and one possible alternative is a bioenergy production system. Biomass, which originates from plants or plant-based materials, has long been investigated as a substitute for fossil fuels.2 Plants or plant-based materials consume CO2 during photosynthesis, then CO2 is released into the atmosphere when these materials are used in energy production (hereinafter named bioenergy production).3 In other words, bioenergy production is deemed CO2 neutral as it results in net zero or minimum increase of CO2 in the atmosphere.3 Following this, Möllersten et al. highlighted the potential of integrating bioenergy production with carbon capture and storage,4 or in short a bioenergy with carbon capture and storage (BECCS) system. Capturing CO2 from biomass burning and storing in a secure geographical reservoir would © 2017 American Chemical Society

Received: February 24, 2017 Revised: April 18, 2017 Published: April 21, 2017 5241

DOI: 10.1021/acssuschemeng.7b00589 ACS Sustainable Chem. Eng. 2017, 5, 5241−5252

Research Article

ACS Sustainable Chemistry & Engineering performance.11 The main reason for ILs to be considered as alternatives to amine-based CO2 capture agents is due their nonvolatile, noncorrosive, nonflammable nature while having high thermal stability and recyclability. These properties make them reduce the energy consumption in CO2 capture processes. Wappel et al. showed that energy demand for carbon capture process using ILs is lower than the process using 30 wt % of MEA solvent.12 Various families of ILs like imidazolium, pyridinium, ammonium, and pyrrolidinium as pure state and binary mixtures have been tested in extensively in recent years in order to search for the most potential candidate for CO2 capture performance.13,14 Improved CO2 absorption performance data have been reported on longer and branched alkyl chains or ether linkages on cation cores or fluorination on cation and anion cores of ILs, thus providing a motive for further research on testing the performance of ILs in CO2 capture processes.15 All of the above-mentioned works indicate that ILs are promising replacements for conventional carbon capture solvents. To install a carbon capture system with an IL-based solvent in a bioenergy production plant, a suitable IL has to be identified to suit the process. However, it is time-consuming and costly to identify the optimal IL solvent by testing about a million possible unique combinations of cations and anions that make up pure ILs10 merely through experiments. To accelerate the identification process, a computer-aided molecular design (CAMD) technique can be implemented to solve the IL design problem.16 A molecule or molecular structure can be identified through CAMD to match a specified set of target properties using this technique.17 Presently, different works have been done to design ILs for specific purposes using the CAMD technique, such as distillation,18 azeotropic separation process,19 and carbon capture.20 However, these works focus on designing ILs at fixed operating conditions, which limits the potential of ILs since they exhibit different performance at varied operating conditions. The product qualities and quantities were shown to be affected when IL-based solvents are used at different temperature.21 Therefore, the performance of ILs determined at fixed conditions does not always correspond to the optimal performamnce within a wide range of feasible temperatures and pressures for a process. Hence, it is essential to identify the optimal IL by considering the performance of ILs within a wide range of feasible temperatures and pressures during the product design stage. Also, the additional utilities required by a carbon capture system can be supplied by a bioenergy system, but these parasitic loads have to be overcome through either an external source supply or retrofitting a bioenergy source. Thus far, there is no reported work presenting integration of BECCS and an IL-based solvent design, considering the subsequent effects in the design problem. The contribution of this work is the introduction of a systematic approach to design ILs for the BECCS purpose, integrating CAMD for IL design and process design, taking the operating conditions and utility requirements of a carbon capture system into consideration. Futhermore, the proposed approach offers a simple graphical tool to decide how to retrofit a bioenergy system accordingly.

production and determine its optimal operating conditions simultaneously from a given set of preselected cation cores, anions, and organic functional groups to meet the target properties and constraints. The optimal IL solvent should be designed based on a design objective, and it must satisfy all target properties and constraints. The approach should also consider the effect of a carbon capture system on the BECCS system and assist in making decisions about retrofitting the entire process to ensure the process is efficient and yet environmental friendly. A generic formulation to design an IL solvent for an existing BECCS process is shown in this section. First, all related information on a bioenergy system is collected. Following that, the amount of CO2 released from bioenergy production and its conditions are determined, followed by identifying the total amount of CO2 captured. This information will then be used to determine final IL that suits the process based on a preset objective, using a previously developed CAMD method.22 Once the IL solvent is identified, the required utilities for the carbon capture system will be calculated, followed by retrofitting the bioenergy system to accommodate the additional utilities required by the carbon capture system. Figure 1 shows the procedure of designing ILs for the BECCS scheme, and it is followed by a detailed description of each step.

Figure 1. Systematic approach to design an IL for a BECCS system.

Step 1: Input−output (IO) modeling is used to represent the existing bioenergy system with CO2 output to determine the conditions and flow rate of the CO2 output. In this system, each component unit is described using only key mass or energy balances and expressed using an IO model in its matrix form, as described by eq 1. (1) Aw = z

■

FORMULATION TO DESIGN AN IONIC LIQUID FOR BIOENERGY WITH CARBON CAPTURE The overall problem to be addressed is stated as follows: Develop a systematic approach to design an optimal IL solvent for the BECCS purpose based on a bioenergy system 5242


Research Article

ACS Sustainable Chemistry & Engineering

Figure 2. Discretization of operating temperature and pressure ranges.

conditions of the carbon capture process are included as part of the optimization objective because operating conditions affect the performance of the IL and process overall. Disjunctive programming is used in this approach to reformulate continuous variables of temperature and pressure as piecewise functions over discrete domains.23 Figure 2 describes how disjunctive programming functions in a simple manner, where operating temperatures and pressures are discretized with each range represented by its midpoint. Equations 4−7 describe the discretization function using disjunctive programming.

where A is the process matrix containing coefficient ratios of mass or energy balances in the system, w is the component unit capacity vector, and z is the final output vector. Following that, a model developed for simultaneous exogenous specifications in the component unit capacity and final output of a given system is used. Equation 2 is obtained by rearrangement of eq 1. ⎡ 0 A′ ⎤⎡ z″ ⎤ ⎡ B′ I ⎤⎡ w″⎤ ⎥⎢ ⎥ ⎥⎢ ⎥ = ⎢ ⎢ ⎣−I A″⎦⎣ w′⎦ ⎣ B″ 0 ⎦⎣ z′ ⎦

(2)

For an l × l matrix (i.e., a square matrix representing a system with l component units with corresponding l main product streams), p is the set of final output (e.g., product streams) with exogenous specifications (z1, z2,..., zp). The remaining (l − p) set is the exogenous specification of the capacity vector that contains the component units with reduced capacity (wp+1, wp+2,..., wl). Here, A′ is the p × p matrix that contains the elements from the first p rows and first k columns in matrix A, and A″ is the (l − p) × p matrix containing the elements from the first (l − p) rows and the first p columns in matrix A. Also, B′ is the p × (l − p) matrix containing the elements from the first p rows and the last (l − p) columns in matrix (−A), and B″ is the (l − p) × (l − p) matrix containing the elements from the first (l − p) rows and (l − p) columns in matrix (−A). Meanwhile, w′ is the p-element column vector containing w1 to wp, which are the endogenous capacities of the component units, and w″ is the (l − p) element column vector containing wp+1 to wl, which are the component units with exogenously specified capacity. Here, z′ is the k-element column vector containing elements z1 to zp, which are the exogenously defined final outputs. Lastly, z″ is the (l − p) element column vector containing elements zp+1 to zl, which are the endogenous final output streams. The total amount of CO2 captured from the bioenergy system is determined then, as this is used for further calculation later. The amount of CO2 captured should meet the emission standard preset by an authorized organization, if it is available, and at the same time fulfills the production demand. Step 2: The design objective and all necessary target properties for the IL design problem are identified according to their actual application, in this case for the BECCS purpose. The optimization objective is written, in general, as shown by eq 3. maximize fg , h

∀ g, h

T gchosen = T gchosen − 1 Ig + Tg(1 − Ig )

(4)

(T L − T gswitch)Ig < T − T gswitch ≤ (T U − T gswitch)(1 − Ig ) (5)

Phchosen = Phchosen − 1 Ih + Ph(1 − Ih)

(6)

(P L − Phswitch)Ih < P − Phswitch ≤ (PU − Phswitch)(1 − Ih) (7) L

U

Here, T and T are lower and upper bounds to any feasible operating temperature, PL and PU are lower and upper bounds for operating pressure, Tgchosen and Phchosen are chosen system temperature and chosen system pressure, Tg and Ph are temperature range g and pressure range h, Tgswitch is boundary temperature between temperature ranges, and Phswitch is boundary pressure between pressure ranges. By introducing this objective function, the model will compare CO2 solubility of all possible ILs between temperature ranges g and g + 1 in eqs 4 and 5 and also between pressure ranges h and h + 1 in eqs 6 and 7. The IL, with the corresponding temperature and pressure ranges, that has the highest amount of absorbed CO2 is chosen, and these will then be compared with the next temperature and pressure ranges. Taking the first two temperature ranges as examples, the first range is T0chosen by default, and this is compared with T1. If an IL in T1 is determined to be the optimal IL among all combinations in both temperature ranges, T1 is chosen as T1chosen according to eq 4. This result will then be compared with all possible ILs in T2. The comparisons are carried on for all predefined temperature ranges up to TU, and a similar procedure is then carried out for all pressure ranges. The upper or lower limit for chosen properties is kept as constraints in solving the design problem, as expressed in eq 8, where τ is the property, and τmin and τmax are the lower and upper bound of these properties.

(3)

where fg,h is the objective of the design problem in temperature range g and pressure range h. Equation 3 shows that the objective is maximized, and maximizing or minimizing the objective is dependent on the objective set. The operating 5243


Research Article


Figure 3. Existing palm oil-based BTS.

τ min ≤ τ ≤ τ max

∑ xi = 1

(8)

∑ yi = 1 i

(10)

i

Step 3: For all properties considered, appropriate prediction models that estimate the target properties of pure ILs are identified in this step. Chong et al. presented a summary of key properties related to pure ILs design,24 for example, CO2 solubility and viscosity. The choices of property models are important because they will affect the accuracy of the presented approach. Step 4: Appropriate molecular building blocks are determined here, including organic groups, cations, and anions to form a complete IL. These molecular building blocks are chosen such that ILs built using these blocks provide properties similar to the conventional products for carbon capture application. Then, data of all target properties for chosen molecular building blocks must be collected for calculation in the next step. Step 5: The optimal IL for the BECCS scheme is designed in this step based on the main objective and property boundaries preset in Step 2. Apart from design objective and property constraints, process and structural constraints are defined as well. In the carbon capture solvent design problem, both liquid and gas phases are involved. Summation of mole fractions for both phases is included as constraints to ensure that vapor fraction, yi, and liquid fractions, xi, always sum to unity, as given by eqs 9 and 10.

The molecular structure of the designed IL must consist of at least two building blocks, as shown in eq 11. Equation 12 indicates that the final IL molecular structure must not have any free bond.

∑ vk ≥ 2

(11)

k

∑ nkvk − 2(∑ nk − 1) = 0 k

k

(12)

In eqs 11 and 12, vk is the number of group k, and nk is the available free bond of group k. In a pure IL, only one cation and one anion should occur, and therefore, eqs 13 and 14 are added to enforce that only one pair of cations and anions will be selected.

∑ αm = 1 m

∑ βn = 1 n

(13)

(14)

In both equations, αm and βn are the binary variables representing each cation m and each anion n, respectively. Step 6: Once the optimal IL is identified, the total utility required by the carbon capture system must be determined to set targets for the bioenergy system.

(9) 5244


Research Article

ACS Sustainable Chemistry & Engineering Table 1. Mass and Energy Balance Data for Technologies in Palm-Based BTSa AD POME (kg h−1) biogas (kg h−1) biomethane (kg h−1) power (kW) high temp. flue gas (kg h−1) released flue gas (kg h−1) return water (kg h−1) cooling water (kg h−1) chilled water (kg h−1) EFB (kg h−1) PMF (kg h−1) dried biomass (kg h−1) HPS (kg h−1) MPS (kg h−1) LPS (kg h−1)

−55,500.000 319.130

DR POME (kg h−1) biogas (kg h−1) biomethane (kg h−1) power (kW) high temp. flue gas (kg h−1) released flue gas (kg h−1) return water (kg h−1) cooling water (kg h−1) chilled water (kg h−1) EFB (kg h−1) PMF (kg h−1) dried biomass (kg h−1) HPS (kg h−1) MPS (kg h−1) LPS (kg h−1)

MM1

MM2

−159.565 157.965 −47.870

−159.565 157.965 −47.870

CT

MCH

−6.310

−57,218.000 56,218.000

GT1

GT2

GT3

−105.310 416.300 526.570

−105.310 416.300 526.570

−105.310 416.300 526.570

HST1

HST2

453.220

453.220

HST3

453.220

HRSG

−1579.710 1579.710

32,558.010

−1671.840

−39,268.050

1671.840

−11,950.800 39,268.050

MST1

161.600

MST2

161.600

−279.000 1000.000

−16,875.000 −9245.300 11,950.800 −13,646.630 13,646.630 −12,814.880

−13,646.630 13,646.630

−13,646.630 13,646.630

WTB1

−20,469.945 20,469.945

−20,469.945 20,469.945

final output −55,500.000 0.000 0.000 2781.840 0.000 34,137.720 −57,218.000 14,999.110 1000.000 −16,875.000 −9245.300 0.000 0.000 0.000 28,125.010

a

AD, anaerobic digester; MM, membrane separator; GT, gas turbine; HRSG, heat recovery steam generator; WTB, water tube boiler. DR, dryer; CT, cooling tower; MCH, mechanical chiller; HST, high pressure steam turbine; MST, medium pressure steam turbine.

produce high pressure steam for heat and power generation. On the other hand, palm oil mill eff luent (POME) is treated in an anaerobic digester, resulting in production of biogas. The biogas (65% methane and 35% CO2) produced is then purified and sent to gas turbines for power generation. Flue gas released from BTS contains a high amount of CO2, and the purpose of implementing BECCS is to separate CO2 from flue gas as a byproduct and reduce CO2 emission of BTS. In this respect, the objective of this case study is to design a suitable IL solvent for postcombustion carbon capture within BECCS scheme. The method of CO2 storage is not the focus of this work, and hence, it is not considered in this case study. First, all related information on the bioenergy system is collected. Table 1 shows the mass and energy balances of each technology within the operating BTS, where positive and negative values denote outputs and inputs, respectively. These values are obtained from the design stage presented in Andiappan et al.25

Step 7: After the required utilities are determined, it is important to evaluate whether the bioenergy system is capable of meeting the required utilities. When the utility of the carbon capture system can be supplied fully by the bioenergy system without affecting the desired energy production, the BECCS system is self-sustainable. On the other hand, if utility requirements of the carbon capture system are not fulfilled by the bioenergy system, the bioenergy system has to be retrofitted to meet the requirement. Different approaches can be applied to reduce the utilities consumption of the carbon capture system. For example, heat integration can be done to reduce the heating and cooling utilities and at the same time reduce operating cost. Another option is to purchase energy from an external source for extra cost without retrofitting the existing bioenergy system.

■

CASE STUDY

■

A case study is presented here to demonstrate the proposed approach. In this case study, the BECCS scheme will be implemented in an existing palm-based biomass trigeneration system (BTS). Figure 3 shows the schematic diagram of an existing palm-based BTS, which utilizes several types of palmbased biomass as feedstock to produce cooling, heating, and power simultaneously. Palm mesocarp f iber (PMF) and empty f ruit bunches (EFBs) are combusted in water tube boilers to

STEP 1: IDENTIFY THE OUTPUTS OF BIOENERGY SYSTEM AND TOTAL CARBON CAPTURED According to Figure 1, the outputs from the bioenergy system are identified and decide the total carbon captured in this step. In this case study, the existing mass and energy balances of BTS are represented using the IO modeling approach. On the basis 5245


Research Article


Figure 4. IO model for BTS.

of the IO modeling approach, the process matrix A consists of the first 15 data rows and first 16 data columns, as shown in Table 1. Meanwhile, the last column constitutes the final output vector z. Each column in matrix A is considered a process vector, wherein key mass or energy balances are presented as ratios and are assumed to be scale-invariant. Following this, the final IO model for BTS is given in Figure 4. As shown, the amount of flue gas produced is 34,138 kg h−1 at an average temperature of 873.15 K. The flue gas stream consists of CO2 and water vapor, and their respective compositions are 72 and 18 wt %. The carbon captured should be specified to comply with a CO2 emission standard. However, there is no CO2 emission standard for industry in Malaysia;26 therefore, 90% of carbon captured is used in this case study for illustration purposes.

Table 2. Compositions of Flue Gas from BTS upper bound

1.0 0

2.0 0.1

■

STEP 3: SELECT PREDICTION MODELS FOR ALL TARGET PROPERTIES Appropriate prediction models are selected in this step to predict the properties of the designed IL. Equations 16−20 are prediction models to estimate density30 and viscosity.31 CO2 solubility of the pure IL is estimated through vapor liquid phase equilibrium as shown in eqs 21 and 22. M ρ= chosen NV (a + bT g + cPhchosen) (16)

STEP 2: IDENTIFY DESIGN OBJECTIVE, TARGET PROPERTIES, AND PROPERTY CONSTRAINTS In this step, the design objective and all influential properties related to the process must be identified. Since the IL is designed for carbon capture purposes, the CO2 solubility (S) of the IL is set as the design objective. It is maximized in this case study; therefore, it is translated into eq 15. ∀ g, h

lower bound

Viscosity of a designed IL solvent should be as low as possible to minimize the pumping power required to circulate it within the process. Therefore, the target range of viscosity is set to be below 0.1 Pa s.29

■

max Sg , h

property density, ρ (g cm−3) viscosity, μ (Pa s)

V=

∑ vkVk

(17)

k

(15)

ln

where Sg,h stands for the solubility of CO2 in selected a IL in temperature range g and pressure range h. For illustration purposes, only the effect of temperature is studied in this case study. The operating pressure is fixed at 0.7 MPa (i.e., Phchosen = 0.7 MPa), while the operating temperature is set as variable, ranging from 323.15 to 373.15 K. The operating temperature is broken down into five ranges, and each temperature range is represented by its midpoint, as illustrated in Figure 5. Nevertheless, this approach can be used to study the effect of pressure or both the temperature and pressure, by including eq 6 and 7.22

Bμ 1000 μ = Aμ + chosen ρM Tg

Aμ =

(18)

∑ vkAμ,k

(19)

k

Bμ =

∑ vkBμ,k

(20)

k chosen yP φi(T , P , yi ) = xiγiPi S i h

(21)

S = xCO2 /x IL

(22) −3

In eq 16, ρ is the IL density in g cm , M is the IL molecular weight in g mol−1, N is the Avogadro constant (given as 0.6022), and V is the molecular volume of IL in Å3. Coefficients a, b, and c were estimated as 0.8005, 6.652 × 10−4 K−1, and −5.919 × 10−4 MPa−1, respectively. Equation 17 shows that the molecular volume of the IL is equivalent to the sum of the molecular volume of each group k (Vk) that occurs in the structure. In eq 18, μ is the IL viscosity in Pa s, and Aμ,k and Bμ,k in eqs 19 and 20 are contributions of group k to parameters Aμ and Bμ. The mole fraction of CO2 in the liquid phase at equilibrium is obtained using eq 21. In this equation, x and yi are the mole fractions of component i in liquid and gas phases, respectively, γi is the activity coefficient determined using the original UNIFAC model, PiS is the saturated vapor pressure of

Figure 5. Midpoints for all temperature ranges considered in the example.

Two other properties are considered in this case study, namely, density (ρ) and viscosity (μ). Density is the basic transport property, and viscosity affects the pumping power requirement and operating cost.27 The target ranges for these properties are included in Table 2. The designed IL solvent should possess similar properties as a conventional carbon capture solvent, i.e., ethanolamines. Hence, the ranges of target properties are set according properties of ethanolamines.28 5246


Research Article

ACS Sustainable Chemistry & Engineering component i, and φi(T,P,yi) is the gas-phase fugacity coefficient. The UNIFAC model consists of a combinatorial contribution, ln γiC, which is essentially due to the differences in size and shape of the molecules and a residual contribution, ln γiR, which is due to energetic interactions,32 as described by eq 23. The details of the UNIFAC model are widely available and, hence, are not further explained in this work. ln γi = ln γi C + ln γi R

∀i

Table 4. Free Bond Number, Molecular Weights, and Data for Target Properties of All Molecular Building Blocks

(23)

Group parameters and binary interaction parameters of IL for the UNIFAC model presented by Lei et al. have been used to predict the activity coefficient in this work.33 ILs are decomposed into a main skeleton of cations and anions, with the remaining alkyl chain connected to the cation broken down into the respective organic function group according to Lei et al.33 In order to apply the UNIFAC model, the cation core (i.e., imidazole ring) is grouped with the anion, while R1 and R2 alkyl chains are broken down into respective organic groups. Equation 24 is an equation of state proposed by Span and Wagner,34 which is used to determine the gas-phase fugacity coefficient. Equation 25 is the extrapolated Antoine equation that is used to calculate the saturated vapor pressure of CO2.35 ln φi = ϕr + δϕδr − ln(1 + δϕδr )

ln PiS = Ai −

Bi T gchosen + Ci

∀i

1 2 3 4

CH3 CH2 [Mim]+ [Im]+

anions

5 6 7

[BF4]− [PF6]− [Cl]−

cation cores

15.03 14.03 82.10 67.07 86.80 144.96 35.45

35 28 119 79 73 107 47

−0.74 −0.63 7.30 8.04 −18.08 −20.49 −27.63

250.0 250.4 1507.1 1257.1 1192.4 2099.8 5457.7

ρ=

(27)

= 0.7

(29)

M 7

0.6022(1.0151) ∑k = 1 vkVk

(30)

7 ⎛ 7 ∑k = 1 vkBμ , k ⎞ ρM ⎜ ⎟ ≤ 0.1 μ= exp⎜∑ vkAμ , k + ⎟ chosen 1000 T ⎝k=1 ⎠ g

xCO2, g =

(31)

0.7y1φ1(T , P , y1) γ1P1S

(32)

ln φi = ϕr + δϕδr − ln(1 + δϕδr )

ln PiS = Ai −

Bi chosen Tg +

C

ln γi = ln γi + ln γi

Table 3. Organic Function Groups, Cation Cores, and Anions Considered organic groups

Bk,μ

1 2 1 2 0 0 0

(28)

STEP 4: SELECT MOLECULAR BUILDING BLOCKS AND COLLECT DATA FOR ALL TARGET PROPERTIES Suitable molecular building blocks are identified in this step. Table 3 shows the molecular fragments that are available for this design problem. These fragments are chosen because they are among the most widely studied cations and anions.36

groups

Ak,μ

1 2 3 4 5 6 7

Phchosen

■

k

Vk

CH3 CH2 [Mim]+ [Im]+ [BF4]− [PF6]− [Cl]−

(T L − T gswitch)Ig < T − T gswitch ≤ (T U − T gswitch)(1 − Ig )

In eq 24, δ = ρ/ρc is the reduced density, while Ai, Bi, and Ci in eq 25 are coefficients of component i. ILs have extremely low vapor pressure;; hence, it is assumed to be negligible in this work. For the same reason, IL is assumed to be absent in vapor phase (i.e., yIL = 0). Both eqs 24 and 25 are employed by Lei et al. to develop the UNIFAC model specifically for ILs.33 Once the mole fraction of CO2 in the liquid phase at equilibrium is determined, eq 22 is used to determine the CO2 solubility of ILs, S in terms of kmole CO2 per kmole IL.

type

molecular weight (g mol−1)

T gchosen = T gchosen − 1 Ig + Tg(1 − Ig )

(25)

groups

nk

STEP 5: DESIGN OPTIMAL IL BASED ON DESIGN OBJECTIVE THAT FULFILLS ALL TARGET PROPERTIES Using data determined in Step 1, formulation of the MINLP model given in the following equations is solved to design an IL solvent for this case study. max Sg = xCO2 /x IL (26)

∀i

k

k

■

(24)

type

building blocks

R

Ci

∀i

(33)

∀i (34)

∀i

(35)

∑ yi = 1

(36)

i

∑ xi = 1

(37)

i

7

∑ vk ≥ 2

(38)

k=1 7

For these selected molecular building blocks, the relevant data of each target property is given in Table 4, as well as the number of free bonds for each functional group. These data are obtained from Gardas and Countiho,30,31 according to the selected prediction models in Step 3.

7

∑ nkvk − 2(∑ nk − 1) = 0 k=1

k=1

(39)

4

∑ αm = 1 m=3

5247

(40) DOI: 10.1021/acssuschemeng.7b00589 ACS Sustainable Chem. Eng. 2017, 5, 5241−5252

Research Article

ACS Sustainable Chemistry & Engineering Table 5. Optimal IL Molecular Design Results IL chosen

chosen system temperature, Tgchosen (K)

[C10MIm][BF4] [C12MIm][BF4] [C9MIm][BF4] [C11MIm][BF4] [C8MIm][BF4] [C10MIm][PF6] [C7MIm][BF4] [C9MIm][PF6] [C6MIm][PF6] [C8MIm][PF6]

328.15 338.15 328.15 338.15 328.15 338.15 328.15 338.15 328.15 338.15

predicted viscosity, μ (PA predicted CO2 solubility, Sg s) 0.08627 0.07121 0.07312 0.06182 0.06193 0.09281 0.05243 0.08133 0.08747 0.07129

0.1320 0.1233 0.1231 0.1163 0.1141 0.1091 0.1049 0.1022 0.09777 0.09509

experimental solubility

rRelative deviation, RD (%)

− − − − 0.111737 − − − 0.108638 −

− − − − 2.15 − − − 9.97 −

Figure 6. Desorption of CO2 from IL solvent. 7

∑ βn = 1 n=5

proposed approach provides the optimal structure of IL for CO2 absorption, based on specifications or performance targets. Predicted solubility can be compared with available experimental data using eq 42. The relative deviations between the available literature experimental solubility data and predicted solubility data are shown in Table 5, showing deviation of less than 10%. Eight out of ten designed ILs are not tested and recorded in the literature to date, and this proposed approach assists in further exploring the possibility of IL application in BECCS. The small difference occurs mainly due to the errors present in the chosen prediction models.

(41)

Equation 26 is the objective function formulated using eqs 16 and 22. Equations 27 and 28 are included to determine the operating temperature range for the carbon capture process, while equation 29 keeps the pressure fixed as it is not part of the objective for this example. Equations 30 and 31 estimate the IL density and viscosity that are formulated based on eqs 16−20. Equations 32−35 are used for CO2 solubility calculation, based on eqs 21−25. The remaining equations are structural constraints formulated from eqs 9−14. The model is solved using LINGO version 14.0, in a computer with an Intel Core i7-4500U processor, and total time required to solve the model is 9 min and 38 s. The optimum IL is determined as [C10MIm][BF4], i.e., 1-decyl-3methylimidazolium tetrafluoroborate, while the optimal operating temperature range is determined as 323.15 to 333.15 K. This results means that [C10mim][BF4] has the highest CO2 solubility among all possible combinations of cations and anions in the identified temperature range. It has a predicted solubility of 0.1320 at 328.15 K and 0.7 MPa, and its viscosity is predicted to be 0.08627 Pa s, which is lower than the upper limit. To date, there is no recorded solubility data for this designed optimum IL in the literature. The developed approach is able to design ILs for the BECCS purpose, even in the absence of experimental data for IL. This is helpful in discovering more ILs for different applications and shortens the time in running experiments. Nine other alternatives are determined by doing integer cuts, along with their respective optimal operating temperature range, and the results are shown in Table 5. As shown, the results obtained by solving this model corresponds to the best ILs that one could find from allowed combinations of anions and cations. This means that the

Relative Deviation |Experimental data − Predicted value| = × 100 Experimental data

(42)

■

STEP 6: DETERMINE UTILITY REQUIRED BY CARBON CAPTURE SYSTEM [C10MIm][BF4] has the highest CO2 solubility according to results in Table 5; hence, it is chosen as the carbon capture solvent for the BECCS scheme in this case study. In this step, the utility required by the carbon capture system using [C10MIm][BF4] as solvent is determined. As presented in Figure 3, power, heating, and cooling utilities are required, and they are calculated using basic energy balance equations. In Step 1, the production of the bioenergy system is identified to be 2782 kW of power, 28,125 kg h−1 of low pressure steam (LPS), and 15,000 kg h−1 of cooling water, corresponding to 24,579.36 kg h−1 of CO2. Here, 90% of this generated CO2 is stated to be captured in the carbon capture system. A basic energy balance is carried out to determine the power consumed to circulate the IL solvent within the system using eqs 43 and 44. 5248


Research Article


qIL =

W=

mCO2M Sg ρ

(43)

qILΔP 3600

(44)

Here, W is the pumping power in kW, qIL is the volumetric flow rate of IL in m3 h−1, mCO2 is the molar flow rate of CO2 in kmol h−1, and ΔP is the pressure difference in kPa. Pumping power required by the carbon capture system is calculated as 193.90 kW. The heating utility is obtained through simulation using Aspen HYSYS version 8.4, which is 10,567 kW.22 The Peng− Robinson fluid package is used to simulate the process because it was shown to be suitable to simulate CO2 desorption from ILs. The flowsheet for CO2 desorption from the IL-based solvent is shown in Figure 6. The total amount of low pressure steam needed by the carbon capture system can be calculated using eq 45. Qh =

FLPSΔh 3600

Figure 7. Comparison of power output from bioenergy system and power required by BECCS system.

(45)

Here, Qh is the heating power in kW, FLPS is the flow rate of LPS in kg h−1, and Δh is the difference in specific enthalpy in kJ kg−1. From the calculation, 14,619.56 kg h−1 of LPS is required to heat the CO2-rich solvent prior to IL solvent regeneration. The cooling utility is determined as 370.28 kW using the same simulation. Qc =

FCW Δh 3600

Figure 8. Comparison of heating output from bioenergy system and heating utility required by BECCS system.

(46)

In eq 46, Qc is the cooling power in kW and FCW is the cooling water flow rate in kg h−1. The total amount of cooling water needed by the carbon capture system is 14,376.27 kg h−1. From the results, it can be concluded that all utilities can be provided by the bioenergy system. However, the production from the bioenergy system is actually for other purposes and does not consider the parasitic loads by the carbon capture system initially. Hence, these extra utilities can either be supplied from external sources or the bioenergy system can be retrofitted to cover them.

■

Figure 9. Comparison of cooling output from bioenergy system and cooling temperature required by BECCS system.

STEP 7: RETROFIT BIOENERGY SYSTEM ACCORDING TO THE REQUIREMENT Once the utilities required by the carbon capture system are calculated, the integrated bioenergy system would require adjustments in output. If the utilities required by the carbon capture system are more than what can be delivered by the bioenergy system, the bioenergy system would require retrofitting. However, once the bioenergy system is retrofitted to produce more utilities, it consequently produces more CO2. Since more CO2 is produced, the carbon capture system would require more utilities to capture CO2. Such a cascading effect is imperative to incorporate when to determine the utility requirement for the carbon capture system. For illustration purposes, power, heating, and cooling utilities are studied, and comparisons are done between the outputs from the bioenergy system and requirements by the entire BECCS system using the [C10MIm][BF4] solvent. Figures 7−9 show the difference between the outputs and requirements in the BECCS system. It can be seen that for all three utilities there are big gaps between the productions and requirements at the beginning. If the

productions are lower than the requirement, the systemrequired utilities are supplied from external sources. When the productions from the bioenergy system are increased, the gaps get narrower until productions meet requirements (marked by black dots on figures). This is the point where the BECCS system is self-sustainable, and no purchasing of utilities is required from other sources. If productions are increased to this point, productions will exceed the requirements of the BECCS system, and they can be sold for profit. According to Figure 7, the power requirement of the carbon capture system can be met when the total power production of the bioenergy system is 2990.4 kW and the corresponding CO2 produced is 26,420.6 kg h−1. When the power production is less than 2990.4 kW, the carbon capture system requires supply from external sources. On the other hand, there is extra power produced from the BECCS system, and it can be sold to a third party when the power production is more than 2990.4 kW. Similarly, the heating utility requirement of the carbon capture 5249



■

ACKNOWLEDGMENTS The financial support from the Ministry of Higher Education, Malaysia, through the LRGS Grant (Project Code: LRGS/ 2013/UKM-UNMC/PT/05) is gratefully acknowledged.

system is met when the output is 42,334 kW as shown in Figure 8, with 51,186 kg h−1. Meanwhile, the cooling utility required by the carbon capture system can be supplied completely by the bioenergy system when the output is 8368 kW and CO2 generated is 59,1543 kg h−1, as shown in Figure 9. With these figures, decision makers can decide how the bioenergy system should be retrofitted. For this case study, if the bioenergy system is retrofitted to fulfill the power requirement only, heating and cooling utilities from external sources are required to run the carbon capture system. If a decision is made to fulfill the required heating utility, the power requirement is fulfilled as well, but the cooling utility should be purchased from an external source. All required utilities are fulfilled when the cooling utility is fulfilled through retrofitting the bioenergy system.

■

BECCS: Bioenergy with carbon capture and storage BTS: Biomass trigeneration system CAMD: Computer-aided molecular design CO2: Carbon dioxide EFB: Empty fruit bunches GC: Group contribution HPS: High pressure steam IL: Ionic liquid IO: Input−output LPS: Low pressure steam MINLP: Mixed integer nonlinear programming MPS: Middle pressure steam PMF: Palm mesocarp fiber POME: Palm oil mill effluent

CONCLUSIONS A systematic approach is presented in this work to design IL as a carbon capture solvent for a BECCS system. The presented approach demonstrates the integration between a bioenergy system and carbon capture system, as a polygeneration system that produces multiple energy sources and CO2 as products. Using the proposed approach, an IL solvent for BECCS purposes can be designed, followed by retrofitting of a bioenergy system to accommodate the utility requirement of a carbon capture system. The IO model is used to determine the outputs from the bioenergy system, and CAMD is applied to solve the IL design problem, identifying the optimal IL and operating conditions of a carbon capture system, based on carbon capture performance and relevant constraints. A simple graphical tool is also presented to assist decision making on retrofitting of a bioenergy system based on productions and requirements within the BECCS system. An illustrative case study is used to demonstrate the application of this proposed approach. In future work, the economic performance of the entire BECCS system can be included to understand the tradeoff between different aspects, including economics, environmental issues, and the feasibility of the process.

Ionic liquids

[MIm]+: Methylimidazolium cation [Im]+: Imidazolium cation [BF4]−: Tetrafluoroborate anion [PF6]−: Hexafluorophosphate anion [Cl]−: Chloride anion Indices

g: Temperature range (g = 1, 2, ..., u) h: Pressure range (h = 1, 2, ..., v) i: Component (i = 1, 2, ..., p) j: Organic functional groups (j = 1, 2, ..., r) k: Groups (k = 1, 2, ..., q) m: Cation groups (m = 1, 2, ..., s) n: Anion groups (n = 1, 2, ..., t) Parameters

a: Coefficient in the model equation for the density A: Process matrix A′: p × p matrix containing elements from first p rows and first k columns in matrix A A″: (l − p) × p matrix containing elements from first (l − p) rows and first p columns in matrix A Ai: Constant for group i in Antoine equation Aμ,k: Contribution of group k to parameter Aμ b: Coefficient in the model equation for the density B′: p × (l − p) matrix containing the elements from the first p rows and the last (l − p) columns in matrix (−A) B″: (l − p) × (l − p) matrix containing the elements from the first (l − p) rows and (l − p) columns in matrix (−A) Bi: Constant for group i in Antoine equation Bμ,k: Contribution of group k to parameter Bμ c: Coefficient in the model equation for the density Ci: Constant for group i in Antoine equation h: Specific enthalpy (kJ kg−1) l: Number of system component units/main product streams N: Avogadro constant nk: Free bond number of group k P: System pressure (MPa) Ph: Pressure range h Phswitch: Boundary pressure between pressure ranges (MPa) PL: Lower bound for system pressure (MPa) PU: Upper bound for system pressure (MPa)

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acssuschemeng.7b00589. Figure S1 shows the process flow diagram for a carbon capture system using IL solvent. Figure S2 shows the decomposition method of an IL which is used for property prediction model in this work. (PDF)

■

NOMENCLATURE

Abbreviations

■

■

Research Article

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Dominic C. Y. Foo: 0000-0002-8185-255X Fadwa T. Eljack: 0000-0003-2869-8933 Mert Atilhan: 0000-0001-8270-7904 Nishanth G. Chemmangattuvalappil: 0000-0002-1501-7441 Notes

The authors declare no competing financial interest. 5250


Research Article


(4) Möllersten, K.; Gao, L.; Yan, J.; Obersteiner, M. Efficient energy systems with CO2 capture and storage from renewable biomass in pulp and paper mills. Renewable Energy 2004, 29 (9), 1583−1598. (5) Möllersten, K.; Yan, J.; Moreira, J. R. Potential market niches for biomass energy with CO2 capture and storageOpportunities for energy supply with negative CO2 emissions. Biomass Bioenergy 2003, 25 (3), 273−285. (6) Zaman, M.; Lee, J. H. Carbon capture from stationary power generation sources: A review of the current status of the technologies. Korean J. Chem. Eng. 2013, 30 (8), 1497−1526. (7) Olajire, A. A. CO2 capture and separation technologies for endof-pipe applications − A review. Energy 2010, 35 (6), 2610−2628. (8) Rai, G.; Kumar, A. Calorimetric elucidation of ionic interactions in room temperature ionic liquid solutions. Clean Technol. Environ. Policy 2014, 16 (8), 1529−1536. (9) Freemantle, M. An Introduction to Ionic Liquids; The Royal Society of Chemistry: Cambridge, 2010. (10) Plechkova, N. V.; Seddon, K. R. Applications of ionic liquids in the chemical industry. Chem. Soc. Rev. 2008, 37 (1), 123−150. (11) Anderson, J. L.; Dixon, J. K.; Brennecke, J. F. Solubility of CO2, CH4, C2H6, C2H4, O2, and N2 in 1-Hexyl-3-methylpyridinium Bis(trifluoromethylsulfonyl)imide: Comparison to Other Ionic Liquids. Acc. Chem. Res. 2007, 40, 1208−1216. (12) Wappel, D.; Gronald, G.; Kalb, R.; Draxler, J. Ionic liquids for post-combustion CO2 absorption. Int. J. Greenhouse Gas Control 2010, 4 (3), 486−494. (13) Karadas, F.; Atilhan, M.; Aparicio, S. Review on the Use of Ionic Liquids (ILs) as Alternative Fluids for CO2 Capture and Natural Gas Sweetening. Energy Fuels 2010, 24 (11), 5817−5828. (14) Ramdin, M.; de Loos, T. W.; Vlugt, T. J. H. State-of-the-Art of CO2 Capture with Ionic Liquids. Ind. Eng. Chem. Res. 2012, 51 (24), 8149−8177. (15) Corvo, M. C.; Sardinha, J.; Casimiro, T.; Marin, G.; Seferin, M.; Einloft, S.; Menezes, S. C.; Dupont, J.; Cabrita, E. J. A Rational Approach to CO2 Capture by Imidazolium Ionic Liquids: Tuning CO 2 Solubility by Cation Alkyl Branching. ChemSusChem 2015, 8 (11), 1935−1946. (16) Eslick, J. C.; Shulda, S. M.; Spencer, P.; Camarda, K. V. Optimization-Based Approaches to Computational Molecular Design. In Molecular Systems Engineering; Adjiman, C. S., Galindo, A., Eds.; Wiley-VCH: Weinheim, Germany, 2010; pp 173−194. (17) Gani, R. Chemical product design: challenges and opportunities. Comput. Chem. Eng. 2004, 28 (12), 2441−2457. (18) Chávez-Islas, L. M.; Vasquez-medrano, R.; Flores-tlacuahuac, A. Optimal Molecular Design of Ionic Liquids for High-Purity Bioethanol Production. Ind. Eng. Chem. Res. 2011, 50 (9), 5153−5168. (19) Roughton, B. C.; Christian, B.; White, J.; Camarda, K. V.; Gani, R. Simultaneous design of ionic liquid entrainers and energy efficient azeotropic separation processes. Comput. Chem. Eng. 2012, 42, 248− 262. (20) Chong, F. K.; Foo, D. C. Y.; Eljack, F. T.; Atilhan, M.; Chemmangattuvalappil, N. G. Ionic liquid design for enhanced carbon dioxide capture by computer-aided molecular design approach. Clean Technol. Environ. Policy 2015, 17 (5), 1301−1312. (21) Díaz, I.; Palomar, J.; Rodríguez, M.; de Riva, J.; Ferro, V.; González, E. J. Ionic liquids as entrainers for the separation of aromatic−aliphatic hydrocarbon mixtures by extractive distillation. Chem. Eng. Res. Des. 2016, 115, 382−393. (22) Chong, F. K.; Foo, D. C. Y.; Eljack, F. T.; Atilhan, M.; Chemmangattuvalappil, N. G. A systematic approach to design taskspecific ionic liquids and their optimal operating conditions. Molecular Systems Design & Engineering 2016, 1, 109−121. (23) El-Halwagi, M. M. Overview of Optimization. In Sustainable Design Through Process Integration: Fundamentals and Applications to Industrial Pollution Prevention, Resource Conservation, and Profitability Enhancement; Elsevier Inc.: Oxford, U.K., 2012; pp 255−286. (24) Chong, F. K.; Eljack, F. T.; Atilhan, M.; Foo, D. C. Y.; Chemmangattuvalappil, N. G. A systematic visual methodology to

T: System temperature (K) Tg: Temperature range g Tgswitch: Boundary pressure between temperature ranges (K) TL: Lower bound for system temperature (K) TU: Upper bound for system temperature (K) Vk: Molecular volume of group k w: Component unit capacity vector w″: (l − p) element column vector containing wp+1 to wl, which are component units with exogenously specified capacity z: Final output vector z′: k-element column vector containing elements z1 to zp, which are exogenously defined final outputs z″: (l − p) element column vector containing elements zp+1 to zl, which are endogenous final output streams Variables

Aμ: Coefficient in the model equation for the viscosity Bμ: Coefficient in the model equation for the viscosity fg,h: Objective of design problem FCW: Cooling water flow rate (kg h−1) FLPS: Low pressure steam flow rate (kg h−1) Ig: Binary variable for temperature range selection Ih: Binary variable for pressure range selection mCO2: Molar flow rate of CO2 (kmol h−1) M: Molecular weight (g mol−1) Phchosen: Chosen system pressure in range h (MPa) PiS: Saturated vapor pressure of component i (MPa) Qc: Cooling duty (kW) Qh: Heating duty (kW) qIL: Volumetric flow rate of IL (m3 h−1) S: CO2 solubility in IL Sg: CO2 solubility in IL in temperature range g (kmol CO2 per kmol IL) Tgchosen: Chosen system temperature in range g (K) V: Molecular volume (Å3) vk: Number of group k W: Pumping power (kW) xi: Mole fraction of component i in liquid phase yi: Mole fraction of component i in gas phase Greek Symbols

αm: Binary variable representing cation m βn: Binary variable representing anion n μ: Viscosity (Pa s) ρ: Density (g cm−3) φi(T,P,yi): Gas-phase fugacity coefficient of component i γi: Activity coefficient of component i γiC: Combinatorial contribution to the activity coefficient of component i γiR: Residual contribution to the activity coefficient of component i τ: Property τmin: Lower bound of property τmax: Upper bound of property

■

REFERENCES

(1) BP Statistical Review of World Energy; British Petroluem, 2013. (2) Kheshgi, H. S.; Prince, R. C.; Marland, G. The Potential of Biomass Fuels in the Context of Global Climate Change: Focus on Transportation Fuels. Annual Review of Energy and the Environment 2000, 25 (1), 199−244. (3) Naik, S. N.; Goud, V. V.; Rout, P. K.; Dalai, A. K. Production of first and second generation biofuels: A comprehensive review. Renewable Sustainable Energy Rev. 2010, 14 (2), 578−597. 5251


Research Article

ACS Sustainable Chemistry & Engineering design ionic liquids and ionic liquid mixtures: Green solvent alternative for carbon capture. Comput. Chem. Eng. 2016, 91, 219−232. (25) Andiappan, V.; Tan, R. R.; Aviso, K. B.; Ng, D. K. S. Synthesis and optimization of biomass-based tri-generation systems with reliability aspects. Energy 2015, 89, 803−818. (26) Environmental Requirements: A Guide For Investors; Department of Environment: Putrajaya, Malaysia, 2010. (27) Kuhlmann, E.; Himmler, S.; Giebelhaus, H.; Wasserscheid, P. Imidazolium dialkylphosphates - a class of versatile, halogen-free and hydrolytically stable ionic liquids. Green Chem. 2007, 9 (3), 233−242. (28) Ethanolamines; The Dow Chemical Company, 2003. (29) Bonhôte, P.; Dias, A.-P.; Papageorgiou, N.; Kalyanasundaram, K.; Grätzel, M. Hydrophobic, Highly Conductive Ambient-Temperature Molten Salts. Inorg. Chem. 1996, 35 (5), 1168−1178. (30) Gardas, R. L.; Coutinho, J. A. P. Extension of the Ye and Shreeve group contribution method for density estimation of ionic liquids in a wide range of temperatures and pressures. Fluid Phase Equilib. 2008, 263 (1), 26−32. (31) Gardas, R. L.; Coutinho, J. A. P. A group contribution method for viscosity estimation of ionic liquids. Fluid Phase Equilib. 2008, 266 (1−2), 195−201. (32) Skjold-Jørgensen, S.; Kolbe, B.; Gmehling, J.; Rasmussen, P. Vapor-Liquid Equilibria by UNIFAC Group Contribution. Revision and Extension. Ind. Eng. Chem. Process Des. Dev. 1979, 18 (4), 714− 722. (33) Lei, Z.; Dai, C.; Wang, W.; Chen, B. UNIFAC model for ionic liquid-CO2 systems. AIChE J. 2014, 60 (2), 716−729. (34) Span, R.; Wagner, W. A New Equation of State for Carbon Dioxide Covering the Fluid Region from the Triple Point Temperature to 1100K at Pressures up to 800 MPa. J. Phys. Chem. Ref. Data 1996, 25 (6), 1509−1596. (35) Shiflett, M. B.; Yokozeki, A. Solubility and diffusivity of hydrofluorocarbons in room-temperature ionic liquids. AIChE J. 2006, 52 (3), 1205−1219. (36) Zhou, F.; Liang, Y.; Liu, W. Ionic liquid lubricants: designed chemistry for engineering applications. Chem. Soc. Rev. 2009, 38 (9), 2590−2599. (37) Gutkowski, K.; Shariati, A.; Peters, C. J. High-pressure phase behavior of the binary ionic liquid system 1-octyl-3-methylimidazolium tetrafluoroborate+carbon dioxide. J. Supercrit. Fluids 2006, 39 (2), 187−191. (38) Shariati, A.; Peters, C. J. High-pressure phase behavior of systems with ionic liquids. J. Supercrit. Fluids 2004, 30 (2), 139−144.

5252


Design of Ionic Liquid as Carbon Capture Solvent ... - ACS Publications

Recommend Documents